Method and Apparatus for Channel Impulse Response Estimation in Gsm Systems

1. A method for estimating channel impulse response (CIR) in a communication system, the method comprising: (a) converting received RF analog signals to obtain baseband digitized signals, sampling the baseband digitized signals over a time according to a symbol period or a bit period, and cross-correlating at least part of the samples and a predetermined set of training sequence symbols; (b) calculating a ratio of a maximum value and a second largest value among squares of a modulus of outputs of the cross-correlations if there are a plurality of peak values among the outputs of the cross-correlation; and (c) if the ratio is larger than a first value, outputting the cross-correlation values as the CIR according to a time step index of which the cross-correlation is the maximum value and an adjacent time step indices.

2. The method of claim 1 further comprising: (d) calculating a number of peak values among the outputs of the cross-correlation.

3. The method of claim 1 , wherein symbols adjacent to the symbol of which the time step index generates the maximum value are a second number of symbols preceding to the symbol of which the time step index generates the maximum value and the second number of symbols succeeding the symbol of which the time step index generates the maximum value.

4. The method of claim 3 , wherein the second number is not less than two.

5. The method of claim 3 , calculating energy of the interval of a predetermined length further comprising calculating squares of the modulus of the outputs of the cross-correlation, adjusting the time step indices of the cross-correlation in turn, and accumulating calculated squares for a third number, and setting the result of the accumulation as the energy of the interval of the predetermined length.

6. The method of claim 1 , wherein the time according to the symbol period or the bit period is a time generated by dividing the symbol period or the bit period by a third integer.

7. The method of claim 6 , wherein the third integer is four.

8. The method of claim 1 , wherein the first value is an integer which is not less than four and not greater than eight.

9. A method for estimating channel impulse response (CIR) in a communication system, the method comprising: (a) converting received RF analog signals to obtain baseband digitized signals, sampling the baseband digitized signals over a time according to a symbol period or a bit period, and cross-correlating at least part of the samples and a predetermined set of training sequence symbols; (b) calculating a ratio of a maximum value and a second largest value among squares of a modulus of outputs of the cross-correlations if there are a plurality of peak values among the outputs of the cross-correlation; and (c) if the ratio is not larger than a first value, calculating energy of an interval of a predetermined length according to the outputs of the cross-correlation, and outputting the cross-correlation value as the CIR according to a beginning time step index of the interval of the predetermined length of which the energy is the maximum and the time step indices following the beginning time step index.

10. The method of claim 9 further comprising: (d) calculating a number of peak values among the outputs of the cross-correlation.

11. The method of claim 9 , wherein symbols adjacent to the symbol of which the time index is the beginning time step index of the interval of the predetermined length of the maximum energy are twice as many as a second number of symbols after the symbol of which the time index is the beginning time step index of the interval of the predetermined length of the maximum energy.

12. The method of claim 11 , wherein the second number is not less than two.

13. The method of claim 9 , wherein calculating the energy of the interval of the predetermined length comprises calculating squares of the modulus of the outputs of the cross-correlation, adjusting the time step indices of the cross-correlation in turn, accumulating calculated squares for a third number, and selling the result of the accumulation as the energy of the interval of the predetermined length.

14. The method of claim 9 , wherein the time according to the symbol period or the bit period is a time generated by dividing the symbol period or the bit period by a third integer.

15. The method of claim 14 , wherein the third integer is four.

16. The method of claim 9 , wherein the first value is an integer which is not less than four, and not greater than eight.

17. A receiver of a communication system, the receiver comprising: a memory for storing a predetermined set of training sequence symbols; a cross-correlator coupled to the memory for performing a predetermined cross-correlation to samples of received symbols and the predetermined set of training sequence symbols; a peak counter for counting a number of peak values among outputs of the cross-correlator; a peak ratio comparator coupled to the peak counter for comparing a ratio of a maximum value and a second largest value among a squares of a modulus of outputs of the cross-correlator with a first value if there are a plurality of peak values among the outputs of the cross-correlator; a module for counting values besides the peak, the module coupled to the peak counter and the peak ratio comparator for outputting the symbol of which the cross-correlator outputs the maximum value according to a time step index and symbols adjacent to the symbol as CIR when there is only one peak value according to the outputs of the cross-correlator or when there are a plurality of peak values among the outputs of the cross-correlator but the ratio of the maximum square of the modulus (power) value and the second largest square of the modulus (power) value among the outputs of the cross-correlator is larger than the first value; and a module for counting N correlation values of a maximum energy, the module coupled to the peak ratio comparator for calculating energy of an interval of a predetermined length according to the outputs of the cross-correlator, and outputting the symbol of which the time step index is the beginning time step index of the interval of the predetermined length of the maximum energy and the adjacent symbols as CIR when there are a plurality of peak values among the outputs of the cross-correlator and the ratio of the maximum square of the modulus (power) value and the second largest square of the modulus (power) value among the outputs of the cross-correlator is not larger than the first value.

18. The receiver of claim 17 , wherein the symbols of which the time step indices are adjacent to the time step index of the symbol of the maximum output of the cross-correlator are a second number of symbols preceding to the symbol of the maximum output of the cross-correlator and the second number of symbols succeeding to the symbol of the maximum output of the cross-correlator.

19. The receiver of claim 18 , wherein the symbols adjacent to the symbol of which the time step index is the beginning time step index of the interval of the predetermined length of the maximum energy are twice as many as a second number of symbols after the symbol of which the time step index is the beginning time step index of the interval of the predetermined length of the maximum energy.

20. The receiver of claim 17 , wherein the module for counting N correlation values of the maximum energy is capable of calculating squares of the modulus of the outputs of the cross-correlator, adjusting the time step indices of the symbols in turn, accumulating calculated squares for a third number, and setting the result of the accumulation as the energy of the interval of the predetermined length.

21. The receiver of claim 17 , wherein the first value is an integer which is not less than four and not greater than eight.

22. The receiver of claim 17 , wherein the second number is not less than two.

23. The receiver of claim 17 , wherein the communication system is a global system for mobile communications (GSM) system.